Dan­ish me­dia take con­trol of AI through the PIN pro­ject

For decades, digitalisation has posed a major challenge for Danish news media. When the internet became part of everyday life, many newspapers made their content freely available online.

This decision would later prove to undermine their business model, as most readers moved to digital platforms. 

When social media became an important distribution channel, the media became dependent on the algorithms, advertising models and analytics tools of the tech giants. 

In both cases, the media responded reactively rather than proactively. However, third time is a charm, and when artificial intelligence began to take its first steps within journalism, an attentive group of researchers had an idea. 

That idea became the research project Platform Intelligence in News (PIN), launched in 2019 and concluded in 2024. The goal was to ensure that AI would boost journalism rather than threaten it. 

"The PIN project set out to help the news industry develop and use AI tools that give them better control of their own data and of how they use artificial intelligence," says Mikkel Flyverbom, Professor at the Department of Management, Society and Communication at CBS and Strategic Lead of the project. 

Mikkel Flyverbom

The PIN project: A proactive strategy for AI in media 

PIN is a collaboration between Copenhagen Business School, JP/Politikens Hus, the Technical University of Denmark and the University of Copenhagen, funded by the Innovation Fund Denmark. The ambition was to develop AI solutions that the media could control themselves. 

Historically, news media have been highly dependent on technologies developed by tech giants like Google and Facebook. These platforms have not only shaped how news is distributed but also created new economic and technological dependencies.  

"In this next wave of digitalisation that is about artificial intelligence, the news industry needs to be better prepared to develop solutions they can control themselves, and that do not further undermine their business model," explains Mikkel Flyverbom. 

The PIN project therefore includes a range of systems and tools designed to help the media industry use artificial intelligence. 

AI helping journalists 

One of the project’s major successes is MAGNA, an AI assistant that supports journalists with research, archive searches and drafting articles. 

"MAGNA is not a replacement for journalists, but a tool that helps them work more efficiently," explains Kasper Lindskow, who served as project manager for PIN while he was a postdoc at CBS and also headed research and innovation at JP/Politikens Hus. He now leads their AI department.  

"It can suggest relevant archive articles, provide an overview of large text volumes of text and help journalists draft article outlines in the tone and style the newspaper is characterised by." 

The PIN project also developed recommender systems to predict readers’ interests and ensure a more balanced news coverage.  

"Some AI models tend to only offer men content on sports and cars – things tied to very stereotypical ideas about men – while showing women content about beauty and housekeeping and things like that," says Mikkel Flyverbom. 

"But when you develop the systems yourself, you can compare them to Google’s and Facebook’s. You can test different models you have developed, see how they work and understand the consequences they have for the news flow readers are exposed to." 

Keeping data in-house 

Thanks to the PIN project, JP/Politiken’s news media have gained several key advantages: 

• They retain ownership of their own data 
• They control how algorithms sort news content, target ads and frame stories 
• They are independent from Silicon Valley’s technologies and business agendas 
• They avoid being steered by algorithms that prioritise, for example, conflict stories outside their own control 

"We are in full control of our data, and our AI systems are designed to support our editorial values rather than being shaped by Silicon Valley standards," says Kasper Lindskow. 

This means journalists do not have to worry about sensitive data ending up with foreign companies or about news flows being skewed by algorithms prioritising values that conflict with those of the newspaper. 

Many media organisations still use major social platforms like Facebook and TikTok to distribute news outside their own ecosystems. But this traffic only accounts for a minority of readers, says Kasper Lindskow. 

"The vast majority of readers are people who actively visit sites like eb.dk or jp.dk. And we are seeing a clear trend: media outlets are increasingly focusing on their own platforms," he says. 

A new kind of tech independence for Danish media 

By developing AI solutions in-house, JP/Politiken has broken away from earlier dependencies on major tech platforms.  

The project has also had international impact and won awards such as WAN-IFRA’s "Best Use of AI in Revenue Strategy" and INMA’s "Global Media Award (press release is in Danish only) ." 

"We have demonstrated that it is possible to use AI to strengthen journalism without sacrificing editorial independence," Kasper Lindskow stresses. 

The outcomes of the PIN project have been shared with the Danish and Nordic media industry through events like Artificial Intelligence in Media and the Nordic AI in Media Summit.  

Parts of the PIN technology have also been made open source, so anyone can use them. 

The PIN project also laid the groundwork for a new AI unit that develops and advises on AI systems across all media in JP/Politiken based on PIN’s technologies and methodologies. Kasper Lindskow now heads this AI unit. 

With the PIN project, Danish media have taken a significant step towards a future where they themselves shape the role AI plays in journalism. The question is whether the rest of the industry will follow.